L. Torrey, J. Shavlik, T. Walker & R. Maclin (2006).
Relational Skill Transfer via Advice Taking. ICML'06 Workshop on Structural Knowledge Transfer for Machine Learning, Pittsburg, PA.

This publication is available in PDF.

Abstract:

We describe a reinforcement learning system that transfers relational skills from a previously learned source task to a related target task. The system uses inductive logic programming to analyze experience in the source task, and transfers rules about when to take actions. The target-task learner accepts these rules through an advice-taking algorithm. Our system also accepts human-provided advice, which can guide the application of transferred skills and provide instruction about new skills in the target task.

Return to the publications of the Univ. of Wisconsin Machine Learning Research Group.